3 ms·
Except that with catchall address, you will inevitably be a collateral from spammers using your domain for fake FROM headers. You'll get a lot of people have au
by varjag 18y ago
Except that with catchall address, you will inevitably be a collateral from spammers using your domain for fake FROM headers. You'll get a lot of people have auto-away messages, spam reply notifications, and "message undeliverable" responses from mail servers in your inbox. And since all of that is legitimate traffic, you can't train Bayesian filter over it.
Been there done that :)
- JimmyL 18y agoMy strategy is to use a catch-all and filters (with Google Apps). I have my catch-all for the domain go to an address like spam@<mydomain>, and then within that account have a filter set up such that all email with a destination that includes my code gets forwarded to my real address. For example, I sign up to HN with the email address hn.rfg@<mydomain>. This gets forwarded to my real address because it's got the rfg string in it. If I start to get spam on that account, I can add a filter for hn.rfg on either my real or spam@<domain> account, depending on which is more convenient, and I know where it came from. Blanket spam to <mydomain> rarely comes through to my email, since the to is unlikely to include the rfg string, and will probably get picked up by the spam filter on the spam@<mydomain> account. Backscatter spam with faked from headers rarely gets to my real account, since the faked address is unlikely to include the rfg string.